A supervised learning method for classifying methylation disorders.

BACKGROUND: DNA methylation is one of the most stable and well-characterized epigenetic alterations in humans. Accordingly, it has already found clinical utility as a molecular biomarker in a variety of disease contexts. Existing methods for clinical diagnosis of methylation-related disorders focus on outlier detection in a small number of CpG sites using standardized cutoffs which differentiate healthy from abnormal methylation levels. The standardized cutoff values used in these methods do not take into account methylation patterns which are known to differ between the sexes and with age. RESULTS: Here we profile genome-wide DNA methylation from blood samples drawn from within a cohort composed of healthy controls of different age and sex alongside patients with Prader-Willi syndrome (PWS), Beckwith-Wiedemann syndrome, Fragile-X syndrome, Angelman syndrome, and Silver-Russell syndrome. We propose a Generalized Additive Model to perform age and sex adjusted outlier analysis of around 700,000 CpG sites throughout the human genome. Utilizing z-scores among the cohort for each site, we deployed an ensemble based machine learning pipeline and achieved a combined prediction accuracy of 0.96 (Binomial 95% Confidence Interval 0.868[Formula: see text]0.995). CONCLUSION: We demonstrate a method for age and sex adjusted outlier detection of differentially methylated loci based on a large cohort of healthy individuals. We present a custom machine learning pipeline utilizing this outlier analysis to classify samples for potential methylation associated congenital disorders. These methods are able to achieve high accuracy when used with machine learning methods to classify abnormal methylation patterns.

Limited Correlation between SARS-CoV-2 Serologic Assays for Identification of High-Titer COVID-19 Convalescent Plasma Using FDA Thresholds.

In August 2020, the Food and Drug Administration (FDA) Emergency Use Authorization (EUA) for COVID-19 convalescent plasma (CCP) specified 12 authorized serologic assays and associated assay-specific cutoff values for the selection of high-titer CCP for use in hospitalized patients. The criteria used for establishing these cutoff values remains unclear. Here, we compare the overall agreement and concordance of five serologic assays included in the August 2020 FDA EUA at both the manufacturer-recommended qualitative cutoff thresholds and at the FDA-indicated thresholds for high-titer CCP, using serum samples collected as part of the CCP Expanded Access Program (EAP). The qualitative positive percent agreement (PPA) across assays ranged from 92.3% to 98.8%. However, the high-titer categorization across assays varied significantly, with the PPA ranging from 26.5% to 82.7%. The Roche anti-NC ECLIA provided the lowest agreement compared to all other assays. Efforts to optimize high-titer cutoffs could reduce, although not eliminate, the discordance across assays. The consequences of using nonstandardized assays are apparent in our study, and the high-titer cutoffs chosen for each assay are not directly comparable to each other. The generalized findings in our study will be relevant to any future use of convalescent plasma for either COVID-19 or future pandemics of newly emerged pathogens. IMPORTANCE COVID-19 convalescent plasma (CCP) was one of the first therapeutic options available for the treatment of SARS-CoV-2 infections and continues to be used selectively for immunosuppressed patients. Given the emergence of novel SARS-CoV-2 variants which are resistant to treatment with available monoclonal antibody (MAb) therapy, CCP remains an important therapeutic consideration. The FDA has released several emergency use authorizations (EUA) that have specified which serological assays can be used for qualification of CCP, as well as assay-specific cutoffs that must be used to identify high-titer CCP. In this study, a cohort of donor CCP was assessed across multiple serological assays which received FDA EUA for qualification of CCP. This study indicates a high degree of discordance across the assays used to qualify CCP for clinical use, which may have precluded the optimal use of CCP, including during clinical trials. This study highlights the need for assay standardization early in the development of serological assays for emerging pathogens.

Identification of Caveolae-Associated Protein 4 Autoantibodies as a Biomarker of Immune-Mediated Rippling Muscle Disease in Adults.

Importance: Immune-mediated rippling muscle disease (iRMD) is a rare myopathy characterized by wavelike muscle contractions (rippling) and percussion- or stretch-induced muscle mounding. A serological biomarker of this disease is lacking. Objective: To describe a novel autoantibody biomarker of iRMD and report associated clinicopathological characteristics. Design, Setting, and Participants: This retrospective cohort study evaluated archived sera from 10 adult patients at tertiary care centers at the Mayo Clinic, Rochester, Minnesota, and Brigham & Women's Hospital, Boston, Massachusetts, who were diagnosed with iRMD by neuromuscular specialists in 2000 and 2021, based on the presence of electrically silent percussion- or stretch-induced muscle rippling and percussion-induced rapid muscle contraction with or without muscle mounding and an autoimmune basis. Sera were evaluated for a common biomarker using phage immunoprecipitation sequencing. Myopathology consistent with iRMD was documented in most patients. The median (range) follow-up was 18 (1-30) months. Exposures: Diagnosis of iRMD. Main Outcomes and Measures: Detection of a common autoantibody in serum of patients sharing similar clinical and myopathological features. Results: Seven male individuals and 3 female individuals with iRMD were identified (median [range] age at onset, 60 [18-76] years). An IgG autoantibody specific for caveolae-associated protein 4 (cavin-4) was identified in serum of patients with iRMD using human proteome phage immunoprecipitation sequencing. Immunoassays using recombinant cavin-4 confirmed cavin-4 IgG seropositivity in 8 of 10 patients with iRMD. Results for healthy and disease-control individuals (n = 241, including myasthenia gravis and immune-mediated myopathies) were cavin-4 IgG seronegative. Six of the 8 individuals with cavin-4 IgG were male, and the median (range) age was 60 (18-76) years. Initial symptoms included rippling of lower limb muscles in 5 of 8 individuals or all limb muscles in 2 of 8 sparing bulbar muscles, fatigue in 9 of 10, mild proximal weakness in 3 of 8, and isolated myalgia in 1 of 8, followed by development of diffuse rippling. All patients had percussion-induced muscle rippling and half had percussion- or stretch-induced muscle mounding. Four of the 10 patients had proximal weakness. Plasma creatine kinase was elevated in all but 1 patient. Six of the 10 patients underwent malignancy screening; cancer was detected prospectively in only 1. Muscle biopsy was performed in 7 of the 8 patients with cavin-4 IgG; 6 of 6 specimens analyzed immunohistochemically revealed a mosaic pattern of sarcolemmal cavin-4 immunoreactivity. Three of 6 patients whose results were seropositive and who received immunotherapy had complete resolution of symptoms, 1 had mild improvement, and 2 had no change. Conclusions and Relevance: The findings indicate that cavin-4 IgG may be the first specific serological autoantibody biomarker identified in iRMD. Depletion of cavin-4 expression in muscle biopsies of patients with iRMD suggests the potential role of this autoantigen in disease pathogenesis.

Targeted Detection of SARS-CoV-2 Nucleocapsid Sequence Variants by Mass Spectrometric Analysis of Tryptic Peptides.

COVID-19 vaccines are becoming more widely available, but accurate and rapid testing remains a crucial tool for slowing the spread of the severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) virus. Although the quantitative reverse transcription-polymerase chain reaction (qRT-PCR) remains the most prevalent testing methodology, numerous tests have been developed that are predicated on detection of the SARS-CoV-2 nucleocapsid protein, including liquid chromatography-tandem mass spectrometry (LC-MS/MS) and immunoassay-based approaches. The continuing emergence of SARS-CoV-2 variants has complicated these approaches, as both qRT-PCR and antigen detection methods can be prone to missing viral variants. In this study, we describe several COVID-19 cases where we were unable to detect the expected peptide targets from clinical nasopharyngeal swabs. Whole genome sequencing revealed that single nucleotide polymorphisms in the gene encoding the viral nucleocapsid protein led to sequence variants that were not monitored in the targeted assay. Minor modifications to the LC-MS/MS method ensured detection of the variants of the target peptide. Additional nucleocapsid variants could be detected by performing the bottom-up proteomic analysis of whole viral genome-sequenced samples. This study demonstrates the importance of considering variants of SARS-CoV-2 in the assay design and highlights the flexibility of mass spectrometry-based approaches to detect variants as they evolve.

SARS-CoV-2 Serologic Assays Dependent on Dual-Antigen Binding Demonstrate Diverging Kinetics Relative to Other Antibody Detection Methods.

Longitudinal studies assessing durability of the anti-severe acute respiratory syndrome coronavirus 2 (anti-SARS-CoV-2) humoral immune response have generated conflicting results. This has been proposed to be due to differences in patient populations, the lack of standardized methodologies, and the use of assays that measure distinct aspects of the humoral response. SARS-CoV-2 antibodies were serially measured in sera from a cohort of 44 well-characterized convalescent plasma donors over 120 days post-COVID-19 symptom onset, utilizing eight assays, which varied according to antigen source, the detected antibody isotype, and the activity measured (i.e., binding, blocking, or neutralizing). While the majority of assays demonstrated a gradual decline in antibody titers over the course of 120 days, the two electrochemiluminescence immunoassay Roche assays (Roche Diagnostics Elecsys anti-SARS-CoV-2 [qualitative, nucleocapsid based] and Roche Diagnostics Elecsys anti-SARS-CoV-2 S [semiquantitative, spike based]), which utilize dual-antigen binding for antibody detection, demonstrated stable and/or increasing antibody titers over the study period. This study is among the first to assess longitudinal, rather than cross-sectional, SARS-CoV-2 antibody profiles among convalescent COVID-19 patients, primarily using commercially available serologic assays with Food and Drug Administration emergency use authorization. We show that SARS-CoV-2 antibody detection is dependent on the serologic method used, which has implications for future assay utilization and clinical value.

Development and validation of IMMUNO-COV™: a high-throughput clinical assay for detecting antibodies that neutralize SARS-CoV-2.

We here describe the development and validation of IMMUNO-COV™, a high-throughput clinical test to quantitatively measure SARS-CoV-2-neutralizing antibodies, the specific subset of anti-SARS-CoV-2 antibodies that block viral infection. The test measures the capacity of serum or purified antibodies to neutralize a recombinant Vesicular Stomatitis Virus (VSV) encoding the SARS-CoV-2 spike glycoprotein. This recombinant virus (VSV-SARS-CoV-2-S-Δ19CT) induces fusion in Vero cell monolayers, which is detected as luciferase signal using a dual split protein (DSP) reporter system. VSV-SARS-CoV-2-S-Δ19CT infection was blocked by monoclonal α-SARS-CoV-2-spike antibodies and by plasma or serum from SARS-CoV-2 convalescing individuals. The assay exhibited 100% specificity in validation tests, and across all tests zero false positives were detected. In blinded analyses of 230 serum samples, only two unexpected results were observed based on available clinical data. We observed a perfect correlation between results from our assay and 80 samples that were also assayed using a commercially available ELISA. To quantify the magnitude of the anti-viral response, we generated a calibration curve by adding stepped concentrations of α-SARS-CoV-2-spike monoclonal antibody to pooled SARS-CoV-2 seronegative serum. Using the calibration curve and a single optimal 1:100 serum test dilution, we reliably measured neutralizing antibody levels in each test sample. Virus neutralization units (VNUs) calculated from the assay correlated closely (p < 0.0001) with PRNT EC50 values determined by plaque reduction neutralization test against a clinical isolate of SARS-CoV-2. Taken together, these results demonstrate that the IMMUNO-COV™ assay accurately quantitates SARS-CoV-2 neutralizing antibodies in human sera and therefore is a potentially valuable addition to the currently available serological tests. The assay can provide vital information for comparing immune responses to the various SARS-CoV-2 vaccines that are currently in development, or for evaluating donor eligibility in convalescent plasma therapy studies.

Gene-Specific Variant Classifier (DPYD-Varifier) to Identify Deleterious Alleles of Dihydropyrimidine Dehydrogenase.

Deleterious variants in dihydropyrimidine dehydrogenase (DPD, DPYD gene) can be highly predictive of clinical toxicity to the widely prescribed chemotherapeutic 5-fluorouracil (5-FU). However, there are very limited data pertaining to the functional consequences of the >450 reported no-synonymous DPYD variants. We developed a DPYD-specific variant classifier (DPYD-Varifier) using machine learning and in vitro functional data for 156 missense DPYD variants. The developed model showed 85% accuracy and outperformed other in silico prediction tools. An examination of feature importance within the model provided additional insight into functional aspects of the DPD protein relevant to 5-FU toxicity. In the absence of clinical data for unstudied variants, prediction tools like DPYD-Varifier have great potential to individualize medicine and improve the clinical decision-making process.

Glycogen Synthase Kinase 3 (GSK-3)-mediated Phosphorylation of Uracil N-Glycosylase 2 (UNG2) Facilitates the Repair of Floxuridine-induced DNA Lesions and Promotes Cell Survival.

Uracil N-glycosylase 2 (UNG2), the nuclear isoform of UNG, catalyzes the removal of uracil or 5-fluorouracil lesions that accumulate in DNA following treatment with the anticancer agents 5-fluorouracil and 5-fluorodeoxyuridine (floxuridine), a 5-fluorouracil metabolite. By repairing these DNA lesions before they can cause cell death, UNG2 promotes cancer cell survival and is therefore critically involved in tumor resistance to these agents. However, the mechanisms by which UNG2 is regulated remain unclear. Several phosphorylation sites within the N-terminal regulatory domain of UNG2 have been identified, although the effects of these modifications on UNG2 function have not been fully explored, nor have the identities of the kinases involved been determined. Here we show that glycogen synthase kinase 3 (GSK-3) interacts with and phosphorylates UNG2 at Thr60 and that Thr60 phosphorylation requires a Ser64 priming phosphorylation event. We also show that mutating Thr60 or Ser64 to Ala increases the half-life of UNG2, reduces the rate of in vitro uracil excision, and slows UNG2 dissociation from chromatin after DNA replication. Using an UNG2-deficient ovarian cancer cell line that is hypersensitive to floxuridine, we show that GSK-3 phosphorylation facilitates UNG2-dependent repair of floxuridine-induced DNA lesions and promotes tumor cell survival following exposure to this agent. These data suggest that GSK-3 regulates UNG2 and promotes DNA damage repair.

Histone H3K27 Trimethylation Modulates 5-Fluorouracil Resistance by Inhibiting PU.1 Binding to the DPYD Promoter.

The antimetabolite 5-fluorouracil (5-FU) is one of the most widely used chemotherapy drugs. Dihydropyrimidine dehydrogenase (DPD) is a major determinant of 5-FU response and toxicity. Although DPYD variants may affect 5-FU metabolism, they do not completely explain the reported variability in DPD function or the resultant differences in treatment response. Here, we report that H3K27 trimethylation (H3K27me3) at the DPYD promoter regulated by Ezh2 and UTX suppresses DPYD expression by inhibiting transcription factor PU.1 binding, leading to increased resistance to 5-FU. Enrichment of H3K27me3 at the DPYD promoter was negatively correlated with both DPYD expression and DPD enzyme activity in peripheral blood specimens from healthy volunteers. Lastly, tumor expression data suggest that DPYD repression by Ezh2 predicts poor survival in 5-FU-treated cancers. Collectively, the findings of the present article suggest that a previously uncharacterized mechanism regulates DPD expression and may contribute to tumor resistance to 5-FU. Cancer Res; 76(21); 6362-73. ©2016 AACR.

microRNAs miR-27a and miR-27b directly regulate liver dihydropyrimidine dehydrogenase expression through two conserved binding sites.

Dihydropyrimidine dehydrogenase (DPD, encoded by DPYD) is the rate-limiting enzyme in the uracil catabolic pathway and has a pivotal role in the pharmacokinetics of the commonly prescribed anticancer drug 5-fluorouracil (5-FU). Deficiency of DPD, whether due to inadequate expression or deleterious variants in DPYD, has been linked to severe toxic responses to 5-FU. Little is known about the mechanisms governing DPD expression in the liver. In this report, we show increased accumulation of RNA-induced silencing complex (RISC) proteins on DPYD mRNA in cells overexpressing the highly homologous microRNAs (miRNA) miR-27a and miR-27b. These miRNAs were shown to repress DPD expression through two conserved recognition sites in DPYD. The IC50 of 5-FU for HCT116 cells overexpressing miR-27a or miR-27b was 4.4 µmol/L (both), significantly lower than that for cells expressing a nontargeting (scramble) control miRNA (14.3 µmol/L; P = 3.3 × 10(-5) and P = 1.5 × 10(-7), respectively). Mouse liver DPD enzyme activity was inversely correlated with expression levels of miR-27a (R(2) = 0.49; P = 0.0012) and miR-27b (R(2) = 0.29; P = 0.022). A common variant in the hairpin loop region of hsa-mir-27a (rs895819) was also shown to be associated with elevated expression of the miR-27a in a panel of cell lines (P = 0.029) and in a transgenic overexpression model (P = 0.0011). Furthermore, rs895819 was associated with reduced DPD enzyme activity (P = 0.028) in a cohort of 40 healthy volunteers. Taken together, these results suggest that miR-27a and miR-27b expression may be pharmacologically relevant modulators of DPD enzyme function in the liver. Furthermore, our data suggest that rs895819 may be a potential risk allele for 5-FU sensitivity.

TRPC3 regulates release of brain-derived neurotrophic factor from human airway smooth muscle.

Exogenous brain-derived neurotrophic factor (BDNF) enhances Ca(2+) signaling and cell proliferation in human airway smooth muscle (ASM), especially with inflammation. Human ASM also expresses BDNF, raising the potential for autocrine/paracrine effects. The mechanisms by which ASM BDNF secretion occurs are not known. Transient receptor potential channels (TRPCs) regulate a variety of intracellular processes including store-operated Ca(2+) entry (SOCE; including in ASM) and secretion of factors such as cytokines. In human ASM, we tested the hypothesis that TRPC3 regulates BDNF secretion. At baseline, intracellular BDNF was present, and BDNF secretion was detectable by enzyme linked immunosorbent assay (ELISA) of cell supernatants or by real-time fluorescence imaging of cells transfected with GFP-BDNF vector. Exposure to the pro-inflammatory cytokine tumor necrosis factor-alpha (TNFα) (20ng/ml, 48h) or a mixture of allergens (ovalbumin, house dust mite, Alternaria, and Aspergillus extracts) significantly enhanced BDNF secretion and increased TRPC3 expression. TRPC3 knockdown (siRNA or inhibitor Pyr3; 10µM) blunted BDNF secretion, and prevented inflammation effects. Chelation of extracellular Ca(2+) (EGTA; 1mM) or intracellular Ca(2+) (BAPTA; 5µM) significantly reduced secreted BDNF, as did the knockdown of SOCE proteins STIM1 and Orai1 or plasma membrane caveolin-1. Functionally, secreted BDNF had autocrine effects suggested by phosphorylation of high-affinity tropomyosin-related kinase TrkB receptor, prevented by chelating extracellular BDNF with chimeric TrkB-Fc. These data emphasize the role of TRPC3 and Ca(2+) influx in the regulation of BDNF secretion by human ASM and the enhancing effects of inflammation. Given the BDNF effects on Ca(2+) and cell proliferation, BDNF secretion may contribute to altered airway structure and function in diseases such as asthma.